AI Hiring Disclosure Requirements in 2026: What to Tell Candidates (With Notice Examples)

Quick answer: AI hiring disclosure requirements depend on where and how a company recruits. In 2026, Ontario requires certain employers to disclose AI screening in public job postings; New York City requires advance notice for covered automated employment decision tools; Illinois has specific AI video-interview rules; and the EU AI Act imposes transparency duties for certain AI interactions. A generic “we use AI” statement is not a universal compliance solution.

Updated 19 October 2026. This is an operational guide for HR teams, not legal advice. Verify jurisdiction, scope, and current rules before using the example wording below.

What is an AI hiring disclosure?

An AI hiring disclosure is information given to applicants or employees about a relevant use of artificial intelligence or automated decision technology in recruitment. It might appear in a job posting, application flow, chatbot, interview invitation or separate notice. The required content and timing vary: one rule may require a short statement in a job ad, another may require advance notice of an assessment tool, and another may require a candidate to consent before an AI-analyzed video interview.

It is important to separate three different questions:

• Notice: Must we tell the person AI will be used, and when?
• Consent: Must we obtain agreement for this particular activity?
• Privacy and decision rights: What additional information, access, safeguards or challenge routes are required under applicable employment and data-protection law?

These are not interchangeable. An AI disclosure does not automatically satisfy a privacy notice, a bias-audit requirement or a consent obligation.

AI hiring notice requirements in 2026: a jurisdiction snapshot

Scope matters. This table is a planning summary, not a substitute for the actual statutes or legal analysis. Official sources are linked at the end of the article.

Ontario: when AI must appear in the job posting

Ontario’s new public-job-posting requirements took effect on 1 January 2026. Under the province’s guidance, a covered employer with at least 25 employees that uses AI to screen, assess or select applicants for a publicly advertised position must disclose that use in the posting. The employer remains responsible if a third-party recruiter performs the relevant activity.

The Ontario guidance says a detailed technical description is not necessary for this specific posting requirement. A plain statement that AI is used for screening, assessment or selection can be sufficient for that element of the rule. The guidance also describes exemptions and how to count employees, so not every job advertisement falls within scope.

Ontario job-posting AI disclosure: sample wording

Example — only where accurate and applicable: “We use artificial intelligence to assist in screening, assessing or selecting applicants for this position. Hiring decisions are managed by our recruitment team.”

The second sentence is an optional factual description, not a statutory formula. Remove it if the workflow does not support the claim. Confirm the exact posting scope and any additional required job-ad statements before publication.

New York City: an AI notice may need to come 10 business days early

New York City’s Local Law 144 applies to certain automated employment decision tools (AEDTs), not every piece of recruiting software with an AI feature. Where it applies, the employer or employment agency must meet the law’s audit and public-summary requirements and provide the required candidate or employee notice at least 10 business days before using the tool.

The notice should identify that an AEDT will be used and the job qualifications or characteristics it will assess. The city’s rules also address information about data sources, data types and retention, as well as requests for an alternative selection process or accommodation. Employers should review the official city rule and FAQ rather than relying on a one-line generic disclosure.

NYC AEDT notice: operational drafting checklist

1. Confirm that the specific product and workflow meet the AEDT definition.
2. Identify the job qualifications and characteristics assessed by the tool.
3. Document the intended first date of use and verify the 10-business-day lead time.
4. Make the required information accessible through a permitted notice channel.
5. Check the current independent bias audit and public summary.
6. Document how candidates can request an alternative selection process or accommodation, as applicable.

Do not simply copy the Ontario sentence into an NYC application. It does not contain the tool-specific information or establish the timing needed for a covered AEDT.

Illinois: distinguish AI video-interview consent from general employment AI notice

Illinois has two relevant regimes that should not be blended into a single rule.

AI-analyzed recorded video interviews

Under the Artificial Intelligence Video Interview Act, an employer requesting applicant-recorded video interviews for Illinois-based positions and using AI to analyse those videos must, before asking applicants to submit them, notify applicants of the AI analysis, explain how the AI works and the general types of characteristics it uses, and obtain consent. A bare “AI may be used” sentence is not enough to address those elements.

A practical pre-interview screen should therefore show the actual tool and assessment information, the required explanation, and a consent control before recording begins. Legal and privacy teams should review recording, sharing and deletion practices separately.

The 2026 Human Rights Act amendment

Public Act 103-0804, effective 1 January 2026, added provisions concerning discriminatory effects from AI in specified employment activities and notice to employees when AI is used for those purposes. Its text expressly uses the word employee in the notice clause. That is a reason to check the statute, any applicable rules and counsel’s interpretation rather than confidently claiming that it imposes the same universal applicant notice in every recruiting scenario.

European Union: Article 50 transparency is already relevant, but not every HR use is the same

The EU AI Act’s Article 50 transparency obligations began applying on 2 August 2026. Article 50(1) places an obligation on providers to ensure certain AI systems intended for direct interaction with natural persons are designed so people are informed they are interacting with AI, unless that is obvious in context. Article 50 also contains other obligations for particular types of AI-generated or manipulated content and specified deployer uses.

For a recruiting team, a candidate-facing chatbot is an obvious place to review the interaction notice and the division of responsibilities between the employer and vendor. A human recruiter privately using AI to draft an email is not automatically the same situation.

EU candidate-facing AI chatbot: plain-language example

“You are chatting with an AI assistant used by our recruiting team. It can answer questions about the application process. For help from a person, contact [recruiting contact].”

This is an illustrative user-experience notice, not a universal Article 50 compliance template. The correct information depends on the system, context, provider/deployer responsibilities, accessibility and any additional privacy requirements.

What about the EU AI Act’s high-risk hiring obligations?

Some AI systems intended to analyse or filter applications or evaluate candidates fall within Annex III, subject to the AI Act’s classification provisions and exceptions. Under the EU’s consolidated implementation timeline, the relevant high-risk rules for Annex III systems are scheduled to apply from 2 December 2027. Article 26(11) includes an information obligation for deployers of in-scope high-risk Annex III systems that make or assist decisions about natural persons.

Do not confuse that later high-risk date with Article 50’s 2026 transparency start. Nor should you assume that every AI feature used by HR is automatically high-risk. For the broader legal picture, read HRYP’s AI Hiring Compliance in 2026 guide.

Which AI hiring disclosure should appear at each candidate touchpoint?

A 30-minute AI candidate-notice readiness check

Before launching another automated recruiting feature, HR can run a focused internal review.

1. List active AI functions (5 minutes). Include sourcing, screening, ranking, interview analysis, chatbots and automated candidate messages.
2. Map candidate and job geography (5 minutes). Check where the role is based, who is assessed, and which laws could apply.
3. Match tools to candidate touchpoints (5 minutes). Determine when the person first encounters the tool or is evaluated by it.
4. Check the actual notice (5 minutes). Verify wording, timing, delivery method, accessibility and any required consent.
5. Assign an owner and preserve evidence (5 minutes). Keep approved copy, screenshots, dates, versions and vendor documentation.
6. Record gaps and next actions (5 minutes). Escalate unclear legal questions before using the feature with real applicants.

If you cannot complete step one confidently, first review how to find shadow AI in recruitment. If your vendor cannot explain what its system assesses, use our AI recruiting vendor due-diligence questions.

Five mistakes that make AI hiring notices unreliable

1. Copying one global disclosure into every market

Ontario job-posting disclosure, NYC AEDT advance notice and Illinois AI video-interview consent solve different legal problems. A universal sentence may be a useful introduction, but it cannot replace the jurisdiction-specific steps.

2. Publishing a notice after the AI has already screened applicants

Timing matters. Build notices into the workflow before the relevant AI assessment or interaction where required, and retain evidence of when the notice was available.

3. Describing an ATS as “administrative” when it ranks candidates

Review the enabled features and their practical effect on candidate progression. An approved software product can still contain a newly enabled AI scoring function.

4. Claiming meaningful human review without checking the interface

If recruiters only see a score and cannot examine the supporting evidence, an override may be more theoretical than real. Review the actual user journey, not just the policy.

5. Assuming the vendor owns every notice obligation

Vendors and employers can have different duties. For example, Ontario’s guidance says the employer remains responsible for the required job-posting disclosure even when a third-party recruiter performs the screening.

Frequently asked questions about AI hiring disclosure

Do employers have to tell candidates they use AI in hiring?

Sometimes, depending on the location, system and use. Ontario has a covered job-posting requirement; NYC has advance notice for covered AEDTs; Illinois requires notice and consent for certain AI-analyzed recorded video interviews; and EU rules address specified AI interactions and high-risk uses on different timelines.

Is a general “we use AI” statement enough?

Not universally. Ontario’s posting requirement can be met with a relatively simple disclosure, but NYC’s covered AEDT notice requires additional tool-related information and advance timing. Illinois video-interview rules require explanation and consent.

When must a NYC candidate receive AI hiring notice?

For an AEDT covered by Local Law 144, the required notice must be provided at least 10 business days before the tool is used. The city also requires a recent bias audit and public summary when the law applies.

Does Ontario require AI disclosure in job postings in 2026?

Yes, for covered employers with at least 25 employees when AI is used to screen, assess or select applicants for a publicly advertised job posting, subject to the provincial rule’s scope and exceptions.

Does the EU AI Act require every AI-assisted email to be labelled?

No blanket rule follows from Article 50. The obligations target specified AI interactions and other defined circumstances. The system’s design, how the candidate interacts with it and the provider/deployer roles all matter.

When do EU high-risk AI hiring notice rules start?

The relevant Annex III high-risk regime, including Article 26 duties for in-scope deployers, is scheduled to apply from 2 December 2027 under the consolidated EU timeline. Separate Article 50 transparency obligations have applied since 2 August 2026.

Can the HRYP assessment generate a legally approved candidate notice?

No. The HRYP AI Hiring Readiness Check is an organisational process assessment. It can help reveal gaps in AI-use visibility, oversight and documentation, but it does not provide legal certification or replace jurisdiction-specific counsel.

The bottom line: map the AI before writing the notice

Good candidate transparency starts with a simple chain: hiring stage → AI function → candidate impact → jurisdiction → notice or consent requirement → evidence → owner. The goal is not to bury applicants in technical language. It is to tell them what they need to know, at the right time, through a channel they can actually use.

For more on how AI interviews differ from ordinary transcription or scheduling, see the AI Interviewer Compliance Checklist.

Official sources and regulatory references

• Government of Ontario — Publicly advertised job posting requirements, including AI disclosure
• NYC Department of Consumer and Worker Protection — Automated Employment Decision Tools
• Illinois General Assembly — Artificial Intelligence Video Interview Act, Section 5
• Illinois General Assembly — Public Act 103-0804
• European Commission AI Act Service Desk — Article 50 transparency obligations
• European Commission AI Act Service Desk — Article 26 deployer obligations
• European Commission AI Act Service Desk — Current enforcement timeline

Editorial note (19 October 2026): This article summarises selected regulatory requirements for HR operations. It is not legal advice, does not establish whether a particular tool is legally an AEDT or high-risk AI system, and does not guarantee that the illustrative notices satisfy local law. Rules, guidance and enforcement can change. Seek qualified advice for actual deployments.